The composite (superposed epoch) analysis technique has been frequently employed to examine a hypothesized link between solar activity and the Earth's atmosphere, often through an investigation of Forbush decrease (Fd) events (sudden high-magnitude decreases in the flux cosmic rays impinging on the upper-atmosphere lasting up to several days). This technique is useful for isolating low-amplitude signals within data where background variability would otherwise obscure detection. The application of composite analyses to investigate the possible impacts of Fd events involves a statistical examination of time-dependent atmospheric responses to Fds often from aerosol and/or cloud datasets. Despite the publication of numerous results within this field, clear conclusions have yet to be drawn and much ambiguity and disagreement still remain. In this paper, we argue that the conflicting findings of composite studies within this field relate to methodological differences in the manner in which the composites have been constructed and analyzed. Working from an example, we show how a composite may be objectively constructed to maximize signal detection, robustly identify statistical significance, and quantify the lower-limit uncertainty related to hypothesis testing. Additionally, we also demonstrate how a seemingly significant false positive may be obtained from non-significant data by minor alterations to methodological approaches.

[1] Svensmark et al. (2009) have recently claimed that strong galactic cosmic ray (GCR) decreases during 'Forbush Decrease (FD) events' are followed by decreases in both the global liquid water cloud fraction (LCF) and other closely correlated atmospheric parameters. To test the validity of these findings we have concentrated on just one property, the MODIS LCF and examined two aspects: 1) The statistical chance that the decrease observed in the LCF is abnormal.2) The likelihood of the observed delay ($5 to 9 days) being physically connected to the FD events. On both counts we conclude that LCF variations are unrelated to FD events: Both the pattern and timing of observed LCF changes are irreconcilable with current theoretical pathways. Additionally, a zonal analysis of LCF variations also offers no support to the claimed relationship, as the observed anomaly is not found to vary latitudinally in conjunction with cosmic ray intensity.

Despite over 35 years of constant satellite-based measurements of cloud, reliable evidence of a long-hypothesized link between changes in solar activity and Earth's cloud cover remains elusive. This work examines evidence of a cosmic ray cloud link from a range of sources, including satellite-based cloud measurements and long-term ground-based climatological measurements. The satellite-based studies can be divided into two categories: (1) monthly to decadal timescale analysis and (2) daily timescale epochsuperpositional (composite) analysis. The latter analyses frequently focus on sudden high-magnitude reductions in the cosmic ray flux known as Forbush Decrease events. At present, two long-term independent global satellite cloud datasets are available (ISCCP and MODIS). Although the differences between them are considerable, neither shows evidence of a solar-cloud link at either long or short timescales. Furthermore, reports of observed correlations between solar activity and cloud over the 1983-1995 period are attributed to the chance agreement between solar changes and artificially induced cloud trends. It is possible that the satellite cloud datasets and analysis methods may simply be too insensitive to detect a small solar signal. Evidence from ground-based studies suggests that some weak but statistically significant cosmic ray-cloud relationships may exist at regional scales, involving mechanisms related to the global electric circuit. However, a poor understanding of these mechanisms and their effects on cloud makes the net impacts of such links uncertain. Regardless of this, it is clear that there is no robust evidence of a widespread link between the cosmic ray flux and clouds. This is anOpen Access article distributed under the terms of Creative Commons Attribution 3.0 Unported License J. Space Weather Space Clim. 2 (2012) A18 A18-p2 B.A. Laken et al.: Cosmic ray climate link and cloud observations A18-p3 J. Space Weather Space Clim. 2 (2012) A18